Stay Ahead of AI Fraud: Spotting and Preventing Scams

Artificial intelligence has only made fraud smarter. According to the FTC, younger people are losing more money to fraud than older people. 44% of 20-29 year olds lost money to fraud in 2024, compared with 24% of 70-79 year olds.

With criminals exploiting AI to carry out their scams, fraud attempts are happening more often and are harder to spot. While misspellings used to be common in scam attempts, generative AI makes it easier for scammers to avoid mistakes and sound legitimate.

If you remain vigilant, there are some tell-tale signs that a communication may be fraud. The first step is knowing what to look for.

AI-Generated Text Scams

Fraudsters will use AI-generated text to pretend to be someone they’re not, whether that’s a love interest or investment advisor. Criminals can even use AI to create fake social media profiles to convince you to send money.

For example, fraudsters will create profiles on dating websites. When you’re coordinating a date with them, they’ll invite you over and say they will order food for you both, asking you to send money to pay for your portion. Once you send over the money, you will never hear from them again.

Criminals use AI-generated text to reach more people faster, embedding chatbots in social media channels, dating sites, and more. Foreign bad actors can even seamlessly translate languages to appear to be local, or generate content for fraudulent cryptocurrency or investment websites so the sites look professional.

AI-Generated Images

When AI first became popular, it was notoriously bad at creating images of humans. There were signs the image was AI-generated, like multiple eyes or a blurred-out nose. However, AI has made huge advances in its image generation. Now, it’s easy to create photo-realistic images of people.

Fraudsters use AI-generated images to perform confidence fraud. Confidence fraud is when someone uses manipulation to gain your trust and convince you to send them money or valuable information. AI-generated photos can help them convince you they are someone they are not, as they can generate lots of images to include in conversations. They can also use the photos to create fake social media profiles.

Impersonation is another tactic fraudsters can more easily employ with AI-generated images. With these images, they can create fake identification documents like driver’s license or credentials. They can also generate images of celebrities or influencers to promote counterfeit products, or even embarrassing photos of a victim to blackmail them into providing payment.

And these fake photos aren’t limited to photos of people. Fraudsters have created images of natural disasters or global conflict to elicit donations to fraudulent charities.

AI-Generated Audio (Vocal Cloning)

If you’ve explored AI-generated audio, you know how realistic it sounds. Just provide it with a sample of the voice you want to emulate (even your own), and it will give you audio that sounds just like that voice, saying whatever you want it to say.

Criminals use this to their advantage in multiple ways. They can impersonate well-known public figures or even a loved one’s voice. With this audio, they can call you and say they are your family member who is in crisis and needs financial assistance – now!

Fraudsters can even impersonate your voice to obtain access to sensitive information or accounts.

AI-Generated Video

If fraudsters can create such realistic photo and audio impersonations, it’s no surprise they can do the same with video. Oftentimes, these videos will be promotional videos endorsing a product like fraudulent investment sites. In cases of confidence fraud, these videos may be sent to try and prove that the person on the other end is “real”.

When fraudsters use videos to impersonate someone, this is called a “deepfake.”  They can hack social media profiles and create deepfakes of that person endorsing a fake product.

They can even use video to engage in real-time video conversations, pretending to be a company executive, law enforcement, or another authority figure.

How to Protect Yourself from AI Fraud

As always, if you receive an urgent or desperate message to send money, it’s important that you slow down. Fraudsters will create a false sense of urgency and prey on your haste to help in a crisis. To protect yourself, do the following:

• Look for subtle imperfections in images and videos like distorted hands and feet, unrealistic teeth and eyes, indistinct or irregular faces, inaccurate shadows, lag time, or watermarks.
• In a video, watch for irregular movements that don’t look typical of the person depicted.
• Pay attention to accessories like glasses or jewelry. If they look unfamiliar, this could be a clue.
• Listen closely for vocal inaccuracies. If it doesn’t sound exactly liked your loved one or boss, maybe it’s not them.
• Verify the identity of the person reaching you by researching the contact of the bank, organization, or person reaching out to you, and call that number directly.

There are also some steps you can take to proactively protect yourself from being a victim of AI-generated fraud:

• Create a secret word with your family to verify their identity
• Limit online content of your image or voice
• Make social media accounts private
• Limit followers to people you know
• Never share sensitive information or money with someone you’ve only met online or over the phone

As always, we are here to help keep you informed on the latest fraud schemes in order to protect you and your information. If you believe you have been a victim of financial fraud, reach out to your local banking center or file a report with the FBI’s Internet Crime Complaint Center at www.ic3.gov.